DESCRIPTION: (Applicant's abstract) The PI's strategy to induce tumor- specific T cells, which is based on years of study in his animal model, has been to immunize patients with autologous tumor that has been combined with BCG or gene-modified to express and allo-antigen. In his murine model, ex vivo direct gene transfer to induce autologous tumor cells to express allo-antigen augments the immunogenicity of the B16BL6 (D5) melanoma. Therapeutic, tumor-specific T cells can be generated from tumor vaccine-draining lymph nodes (TVDLN) following vaccination with gene- modified cells but not unmodified tumor. Adoptive transfer of these T cells to tumor-bearing mice after activation with anti-D3 and IL-2 led to significant reductions in pulmonary metastases. However, vaccinations with unmodified tumor is ineffective at sensitizing T cells with therapeutic activity and fails to protect animals from a subsequent tumor challenge. The applicant has considered the possibility that the host simply fails to recognize this poorly/non-immunogenic tumor. However, examination of LN draining the poorly immunogenic D5 tumor revealed an increased number of cells with the L-selectinLO/- phenotype, suggesting that T cells were responding to vaccination. Following in vitro activation with anti-CD3, L- selectinLO/-cells from D5 vaccines exhibited tumor-specific cytokine secretion that was highly polarized to a type 2 (T2) profile (IL-4 and IL- 10) whereas vaccination with gene modified D5-kD led to type 1 (T1) cytokine release (IFN-gamma). The failure to generate therapeutic T cells from lymph nodes draining the unmodified vaccine is related to the induction of an immune response that is ineffective in mediating tumor regression, not to an inability to recognize tumor antigens. These observations led the PI to formulate the following two hypothesis: 1) the isolation of L-selectinLO/- cells from lymph nodes draining the vaccine site will enrich the tumor-reactive T cells, and 2) the development of a tumor-specific T1 cytokine response is critical to the generation of critical to the generation of therapeutic T cells for adoptive immunotherapy. To test these hypothesis, it is essential that the immune response of patients to vaccination by monitored appropriately. This will require an emphasis on the regulatory cytokines that play a pivotal role in polarizing the immune response. The I will carefully evaluate the baseline immune function of the 30 patients entered onto these trials. He will then evaluate the immune responses to vaccination and evaluate whether a tumor-specific immune response is generated and whether it is polarized to either a T1 or T2 cytokine profile. Since it will be difficult to asses the response of a number of tumor-sensitized T cells within a large population of non-tumor sensitized T cells in the TVDLN, he will enrich for antigen reactive cells by isolating L-selectinLO/-cells, which will be evaluated for tumor-specific and polarized cytokine responses. This will allow him to test whether the hypothesis derived from the animal studies are relevant to human tumor vaccine strategies and thus form a basis to assume that these findings are basic principles that are generally applicable for the treatment of cancer.